Two-Layer Attention Feature Pyramid Network for Small Object Detection

Effective small object detection is crucial in various applications including urban intelligent transportation and pedestrian detection.However,small objects are difficult to detect accurately because they contain less information.Many current methods,particularly those based on Feature Pyramid Network(FPN),address this challenge by leveraging multi-scale feature fusion.However,existing FPN-based methods often suffer from inadequate feature fusion due to varying resolutions across different layers,leading to suboptimal small object detection.To address this problem,we propose the Two-layerAttention Feature Pyramid Network(TA-FPN),featuring two key modules:the Two-layer Attention Module(TAM)and the Small Object Detail Enhancement Module(SODEM).TAM uses the attention module to make the network more focused on the semantic information of the object and fuse it to the lower layer,so that each layer contains similar semantic information,to alleviate the problem of small object information being submerged due to semantic gaps between different layers.At the same time,SODEM is introduced to strengthen the local features of the object,suppress background noise,enhance the information details of the small object,and fuse the enhanced features to other feature layers to ensure that each layer is rich in small object information,to improve small object detection accuracy.Our extensive experiments on challenging datasets such as Microsoft Common Objects inContext(MSCOCO)and Pattern Analysis Statistical Modelling and Computational Learning,Visual Object Classes(PASCAL VOC)demonstrate the validity of the proposedmethod.Experimental results show a significant improvement in small object detection accuracy compared to state-of-theart detectors.

Assessing edge-coupled interdependent network disintegration via rank aggregation and elite enumeration

The disintegration of networks is a widely researched topic with significant applications in fields such as counterterrorism and infectious disease control. While the traditional approaches for achieving network disintegration involve identifying critical sets of nodes or edges, limited research has been carried out on edge-based disintegration strategies. We propose a novel algorithm, i.e., a rank aggregation elite enumeration algorithm based on edge-coupled networks(RAEEC),which aims to implement tiling for edge-coupled networks by finding important sets of edges in the network while balancing effectiveness and efficiency. Our algorithm is based on a two-layer edge-coupled network model with one-to-one links, and utilizes three advanced edge importance metrics to rank the edges separately. A comprehensive ranking of edges is obtained using a rank aggregation approach proposed in this study. The top few edges from the ranking set obtained by RAEEC are then used to generate an enumeration set, which is continuously iteratively updated to identify the set of elite attack edges.We conduct extensive experiments on synthetic networks to evaluate the performance of our proposed method, and the results indicate that RAEEC achieves a satisfactory balance between efficiency and effectiveness. Our approach represents a significant contribution to the field of network disintegration, particularly for edge-based strategies.

Percolation transitions in edge-coupled interdependent networks with directed dependency links

We propose a model of edge-coupled interdependent networks with directed dependency links(EINDDLs)and develop the theoretical analysis framework of this model based on the self-consistent probabilities method.The phase transition behaviors and parameter thresholds of this model under random attacks are analyzed theoretically on both random regular(RR)networks and Erd¨os-Renyi(ER)networks,and computer simulations are performed to verify the results.In this EINDDL model,a fractionβof connectivity links within network B depends on network A and a fraction(1-β)of connectivity links within network A depends on network B.It is found that randomly removing a fraction(1-p)of connectivity links in network A at the initial state,network A exhibits different types of phase transitions(first order,second order and hybrid).Network B is rarely affected by cascading failure whenβis small,and network B will gradually converge from the first-order to the second-order phase transition asβincreases.We present the critical values ofβfor the phase change process of networks A and B,and give the critical values of p andβfor network B at the critical point of collapse.Furthermore,a cascading prevention strategy is proposed.The findings are of great significance for understanding the robustness of EINDDLs.

Cooperative Caching Strategy Based on Two-Layer Caching Model for Remote Sensing Satellite Networks

In Information Centric Networking(ICN)where content is the object of exchange,in-network caching is a unique functional feature with the ability to handle data storage and distribution in remote sensing satellite networks.Setting up cache space at any node enables users to access data nearby,thus relieving the processing pressure on the servers.However,the existing caching strategies still suffer from the lack of global planning of cache contents and low utilization of cache resources due to the lack of fine-grained division of cache contents.To address the issues mentioned,a cooperative caching strategy(CSTL)for remote sensing satellite networks based on a two-layer caching model is proposed.The two-layer caching model is constructed by setting up separate cache spaces in the satellite network and the ground station.Probabilistic caching of popular contents in the region at the ground station to reduce the access delay of users.A content classification method based on hierarchical division is proposed in the satellite network,and differential probabilistic caching is employed for different levels of content.The cached content is also dynamically adjusted by analyzing the subsequent changes in the popularity of the cached content.In the two-layer caching model,ground stations and satellite networks collaboratively cache to achieve global planning of cache contents,rationalize the utilization of cache resources,and reduce the propagation delay of remote sensing data.Simulation results show that the CSTL strategy not only has a high cache hit ratio compared with other caching strategies but also effectively reduces user request delay and server load,which satisfies the timeliness requirement of remote sensing data transmission.

NEAR-INFRARED OPTICAL TOMOGRAPHY IMAGE RECONSTRUCTION APPROACH BASED ON TWO-LAYERED BP NEURAL NETWORK

An image-reconstruction approach for optical tomography is presented,in which a two-layered BP neural network is used to distinguish the tumor location.The inverse problem is solved as optimization problem by Femlab software and Levenberg–Marquardt algorithm.The concept of the average optical coefficient is proposed in this paper,which is helpful to understand the distribution of the scattering photon from tumor.The reconstructive¯µ
s by the trained network is reasonable for showing the changes of photon number transporting inside tumor tissue.It realized the fast reconstruction of tissue optical properties and provided optical OT with a new method.

Identification of Important FPGA Modules Based on Complex Network

The globalization of hardware designs and supply chains,as well as the integration of third-party intellectual property(IP)cores,has led to an increased focus from malicious attackers on computing hardware.However,existing defense or detection approaches often require additional circuitry to perform security verification,and are thus constrained by time and resource limitations.Considering the scale of actual engineering tasks and tight project schedules,it is usually difficult to implement designs for all modules in field programmable gate array(FPGA)circuits.Some studies have pointed out that the failure of key modules tends to cause greater damage to the network.Therefore,under limited conditions,priority protection designs need to be made on key modules to improve protection efficiency.We have conducted research on FPGA designs including single FPGA systems and multi-FPGA systems,to identify key modules in FPGA systems.For the single FPGA designs,considering the topological structure,network characteristics,and directionality of FPGA designs,we propose a node importance evaluationmethod based on the technique for order preference by similarity to an ideal solution(TOPSIS)method.Then,for the multi-FPGA designs,considering the influence of nodes in intra-layer and inter-layers,they are constructed into the interdependent network,and we propose a method based on connection strength to identify the important modules.Finally,we conduct empirical research using actual FPGA designs as examples.The results indicate that compared to other traditional indexes,node importance indexes proposed for different designs can better characterize the importance of nodes.

Opinion consensus incorporating higher-order interactions in individual-collective networks

In the current information society, the dissemination mechanisms and evolution laws of individual or collective opinions and their behaviors are the research hot topics in the field of opinion dynamics. First, in this paper, a two-layer network consisting of an individual-opinion layer and a collective-opinion layer is constructed, and a dissemination model of opinions incorporating higher-order interactions(i.e. OIHOI dissemination model) is proposed. Furthermore, the dynamic equations of opinion dissemination for both individuals and groups are presented. Using Lyapunov's first method,two equilibrium points, including the negative consensus point and positive consensus point, and the dynamic equations obtained for opinion dissemination, are analyzed theoretically. In addition, for individual opinions and collective opinions,some conditions for reaching negative consensus and positive consensus as well as the theoretical expression for the dissemination threshold are put forward. Numerical simulations are carried to verify the feasibility and effectiveness of the proposed theoretical results, as well as the influence of the intra-structure, inter-connections, and higher-order interactions on the dissemination and evolution of individual opinions. The main results are as follows.(i) When the intra-structure of the collective-opinion layer meets certain characteristics, then a negative or positive consensus is easier to reach for individuals.(ii) Both negative consensus and positive consensus perform best in mixed type of inter-connections in the two-layer network.(iii) Higher-order interactions can quickly eliminate differences in individual opinions, thereby enabling individuals to reach consensus faster.

The Invulnerability of Directed Interdependent Networks with Multiple Dependency Relations

The paper aims to study the invulnerability of directed interdependent networks with multiple dependency relations: dependent and supportive. We establish three models and simulate in three network systems to deal with this question. To improve network invulnerability, we’d better avoid dependent relations transmission and add supportive relations symmetrically.

Cooperation influenced by the correlation degree of two-layered complex networks in evolutionary prisoner’s dilemma games

An evolutionary prisoner＇s dilemma game is investigated on two-layered complex networks respectively representing interaction and learning networks in one and two dimensions. A parameter q is introduced to denote the correlation degree between the two-layered networks. Using Monte Carlo simulations we studied the effects of the correlation degree on cooperative behaviour and found that the cooperator density nontrivially changes with q for different payoff parameter values depending on the detailed strategy updating and network dimension. An explanation for the obtained results is provided.

Synchronizability of Two-Layer Cluster Ring Networks

Multilayer network is a frontier direction of network science research. In this paper, the cluster ring network is extended to a two-layer network model, and the inner structures of the cluster blocks are random, small world or scale-free. We study the influence of network scale, the interlayer linking weight and interlayer linking fraction on synchronizability. It is found that the synchronizability of the two-layer cluster ring network decreases with the increase of network size. There is an optimum value of the interlayer linking weight in the two-layer cluster ring network, which makes the synchronizability of the network reach the optimum. When the interlayer linking weight and the interlayer linking fraction are very small, the change of them will affect the synchronizability.

Assessing Air Traffic Management Performance Interdependencies through Bayesian Networks： Preliminary Applications and Results

The performance model proposed by this study represents an innovative approach to deal with performance assessment in ATM （air traffic management）. It is based on Bayesian networks methodology, which presents several advantages but also some drawbacks as highlighted along the paper. We illustrate the main steps required for building the model and present a number of interesting results. The contribution of the paper is two-fold： （1） It presents a new methodological approach to deal with a problem which is of strategic importance for ANSPs （air navigation service providers）; （2） It provides insights on the interdependencies between factors influencing performance. Both results are considered particularly important nowadays, due to the SES （Single European Sky） performance scheme and its related target setting process.

Target Inactivation and Recovery in Two-Layer Networks

We study the target inactivation and recovery in two-layer networks. Five kinds of strategies are chosen to attack the two-layer networks and to recover the activity of the networks by increasing the inter-layer coupling strength. The results show that we can easily control the dying state effectively by a randomly attacked situation. We then investigate the recovery activity of the networks by increasing the inter-layer coupled strength. The optimal values of the inter-layer coupled strengths are found, which could provide a more effective range to recovery activity of complex networks. As the multilayer systems composed of active and inactive elements raise important and interesting problems, our results on the target inactivation and recovery in two-layer networks would be extended to different studies.

Semantic Path Attention Network Based on Heterogeneous Graphs for Natural Language to SQL Task

The natural language to SQL(NL2SQL)task is an emerging research area that aims to transform a natural language with a given database into an SQL query.The earlier approaches were to process the input into a heterogeneous graph.However,previous models failed to distinguish the types of multi-hop connections of the heterogeneous graph,which tended to ignore crucial semantic path information.To this end,a two-layer attention network is presented to focus on essential neighbor nodes and mine enlightening semantic paths for feature encoding.The weighted edge is introduced for schema linking to connect the nodes with semantic similarity.In the decoding phase,a rule-based pruning strategy is offered to refine the generated SQL queries.From the experimental results,the approach is shown to learn a good encoding representation and decode the representation to generate results with practical meaning.

Evolutionary Games in Two-Layer Networks with the Introduction of Dominant Strategy

We study evolutionary games in two-layer networks by introducing the correlation between two layers through the C-dominance or the D-dominance. We assume that individuals play prisoner＇s dilemma game （PDG） in one layer and snowdrift game （SDG） in the other. We explore the dependences of the fraction of the strategy cooperation in different layers on the game parameter and initial conditions. The results on two-layer square lattices show that, when cooperation is the dominant strategy, initial conditions strongly influence cooperation in the PDG layer while have no impact in the SDG layer. Moreover, in contrast to the result for PDG in single-layer square lattices, the parameter regime where cooperation could be maintained expands significantly in the PDG layer. We also investigate the effects of mutation and network topology. We find that different mutation rates do not change the cooperation behaviors. Moreover, similar behaviors on cooperation could be found in two-layer random networks.

基于相依网络的无人机集群通信系统鲁棒性分析

作为地震灾害救援和侦察战场等的主力设备,无人机集群通信系统的鲁棒性是抗毁性的重要指标,本文借助复杂相依网络理论建模无人机集群通信系统,引入相对网络效能比作为通信系统网络整体鲁棒性的指标,准确地反映节点之间的相互依赖关系,为评估通信系统在目标攻击下的鲁棒性提供了一种全新的视角。基于经典M-L模型和动态信息的负载重分配策略,建立相依同构ER-ER和异构ER-BA网络模型,考虑两层网络中相依节点度数的特性进行目标攻击,分析其在攻击下的级联失效动态过程,以及任意两节点的连边概率、平均度数和容量系数等特征参数,对通信网络整体鲁棒性的影响。研究表明,网络鲁棒性和网络相依强度负相关,在级联失效过程中可考虑降低52.09%~72.9%相依强度,以抵抗更严重的通信网络崩塌;对同构网络来说,相依模式的不同,RL、AL较DL整体鲁棒性分别降低了9.25%和15.95%。

城市关联基础设施多层网络韧性评估模型

为了探究城市基础设施系统关联关系对城市韧性的影响,采用理论分析和python模拟相结合的方法研究城市关联基础设施系统在扰动下的韧性变化过程。首先,基于韧性机制和多层网络理论,构建城市关联基础设施网络;然后,探究在功能关联下,城市关联基础设施网络受到3种外界扰动后的级联失效过程,并基于基础设施系统机能曲线法,构建城市关联基础设施网络韧性评估模型;最后,以湖南省中部偏北某县作为研究对象进行实例研究。结果表明:关联网络中电力节点的聚类系数明显高于整体平均水平;通信网络的抵御能力和恢复能力是5类基础设施网络中最强的,韧性水平也最高;随着网络中节点失效概率增大和网络层数的增加,整体关联基础设施网络及其子网络的可靠性逐渐降低;网络效率在介数恢复策略下恢复至较高水平时间较短,说明对此基础设施网络使用介数恢复策略更为合适。

基于相依失效的“空-基-地”编队可靠性分析

针对大规模多域无人设备之间难以进行可靠性分析的问题,根据实际“空-基-地”分布式协同编队特征,提出了一种基于多域无人平台运作规则与层网络间逻辑连接失效关系的空基地集群编队可靠性分析算法。用空中网络、通信网络、地面网络组成的3层轻量化方格拓扑结构代替现实中的集群,在综合考虑随机攻击、蓄意攻击以及区域失效关系、关联失效关系、级联失效关系等依赖关系的条件下设计算法,以网络中最大连通子图为可靠性测度,进行不同编队状态及编队形式下的网络可靠性分析。最后与标准仿真平台进行对比,试验结果验证了算法的可行性、快速性以及资源占用率小等优势。该研究为提升多域无人平台协同编队时的可靠性及相关的防控策略的研究提供了研究基础。

二元双层耦合网络渗流行为分析

本文建立了节点和连边耦合的二元相互依赖网络模型(binary interdependent networks with node and edge coupling,BINNEC).该模型中,网络A节点随机依赖网络B多条边从而形成边依赖群,当边依赖群的失效比例超过网络失效容忍度μ时,导致网络A的节点失效.基于自平衡概率法建立起理论分析框架,通过对3种经典网络结构下的BINNEC网络进行计算机仿真验证,并分析了该模型面对随机攻击下的相变行为和关键阈值.研究结果表明3种网络结构下的BINNEC网络和一元耦合网络一样脆弱,网络的相变行为都是一阶相变.当网络节点的边依赖群规模m越大,网络鲁棒性增强;在网络的边依赖群规模一定时,节点的失效容忍度μ越大网络鲁棒性越强.本文进一步研究在m和μ相同条件下,网络的度分布对BINNEC的影响.研究可以为此类二元耦合网络鲁棒性提供一定的理论依据,同时也为现实网络的安全设计提供指导意义.

负载目标下基于延迟分配的多层级相依网络级联失效模型

在负载具有目标节点的多层级相依网络中,为了避免大规模级联失效的发生、增加网络鲁棒性、提高负载完成效率,提出一种可调参数的级联失效模型。该模型设置了新的初始负载和节点容量,构建了负载延迟分配机制,在负载分配的异质性和均匀性基础上制定了负载重分配规则。解析分析模型的级联失效条件,并通过WS-BA相依网络进行了仿真实验,结果表明,在网络容量一定条件下,合理调节异质性参数和均匀性参数可以提高网络的鲁棒性和效率。

基于博弈论的关联网络攻防博弈分析

根据现实中关联网络复杂的关联特性和网络中重要的评估指标,提出了5种不同的耦合方式,建立了9种关联网络模型;考虑到网络中信息的传输,再分配和级联故障,建立了基于介数虚拟流的级联失效模型;基于博弈论,从复杂网络的角度分析了关键基础设施的攻防问题,分析了多种关联网络的鲁棒性。发现了关联网络中博弈参与者的偏好,为基础设施网络的保护提供了决策支持。